bitmap.c 55.8 KB
Newer Older
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
/*
 * bitmap.c two-level bitmap (C) Peter T. Breuer (ptb@ot.uc3m.es) 2003
 *
 * bitmap_create  - sets up the bitmap structure
 * bitmap_destroy - destroys the bitmap structure
 *
 * additions, Copyright (C) 2003-2004, Paul Clements, SteelEye Technology, Inc.:
 * - added disk storage for bitmap
 * - changes to allow various bitmap chunk sizes
 */

/*
 * Still to do:
 *
 * flush after percent set rather than just time based. (maybe both).
 */

18
#include <linux/blkdev.h>
19 20 21 22 23 24 25 26 27 28
#include <linux/module.h>
#include <linux/errno.h>
#include <linux/slab.h>
#include <linux/init.h>
#include <linux/timer.h>
#include <linux/sched.h>
#include <linux/list.h>
#include <linux/file.h>
#include <linux/mount.h>
#include <linux/buffer_head.h>
29
#include "md.h"
30
#include "bitmap.h"
31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52

/* debug macros */

#define DEBUG 0

#if DEBUG
/* these are for debugging purposes only! */

/* define one and only one of these */
#define INJECT_FAULTS_1 0 /* cause bitmap_alloc_page to fail always */
#define INJECT_FAULTS_2 0 /* cause bitmap file to be kicked when first bit set*/
#define INJECT_FAULTS_3 0 /* treat bitmap file as kicked at init time */
#define INJECT_FAULTS_4 0 /* undef */
#define INJECT_FAULTS_5 0 /* undef */
#define INJECT_FAULTS_6 0

/* if these are defined, the driver will fail! debug only */
#define INJECT_FATAL_FAULT_1 0 /* fail kmalloc, causing bitmap_create to fail */
#define INJECT_FATAL_FAULT_2 0 /* undef */
#define INJECT_FATAL_FAULT_3 0 /* undef */
#endif

53
static inline char *bmname(struct bitmap *bitmap)
54 55 56 57 58 59 60 61 62 63 64
{
	return bitmap->mddev ? mdname(bitmap->mddev) : "mdX";
}

/*
 * just a placeholder - calls kmalloc for bitmap pages
 */
static unsigned char *bitmap_alloc_page(struct bitmap *bitmap)
{
	unsigned char *page;

65
#ifdef INJECT_FAULTS_1
66 67
	page = NULL;
#else
68
	page = kzalloc(PAGE_SIZE, GFP_NOIO);
69 70 71 72
#endif
	if (!page)
		printk("%s: bitmap_alloc_page FAILED\n", bmname(bitmap));
	else
73 74
		pr_debug("%s: bitmap_alloc_page: allocated page at %p\n",
			 bmname(bitmap), page);
75 76 77 78 79 80 81 82
	return page;
}

/*
 * for now just a placeholder -- just calls kfree for bitmap pages
 */
static void bitmap_free_page(struct bitmap *bitmap, unsigned char *page)
{
83
	pr_debug("%s: bitmap_free_page: free page %p\n", bmname(bitmap), page);
84 85 86 87 88 89 90 91 92 93 94 95 96
	kfree(page);
}

/*
 * check a page and, if necessary, allocate it (or hijack it if the alloc fails)
 *
 * 1) check to see if this page is allocated, if it's not then try to alloc
 * 2) if the alloc fails, set the page's hijacked flag so we'll use the
 *    page pointer directly as a counter
 *
 * if we find our page, we increment the page's refcount so that it stays
 * allocated while we're using it
 */
97 98
static int bitmap_checkpage(struct bitmap *bitmap,
			    unsigned long page, int create)
99 100
__releases(bitmap->lock)
__acquires(bitmap->lock)
101 102 103 104
{
	unsigned char *mappage;

	if (page >= bitmap->pages) {
105 106 107 108
		/* This can happen if bitmap_start_sync goes beyond
		 * End-of-device while looking for a whole page.
		 * It is harmless.
		 */
109 110 111 112 113 114 115 116 117 118 119 120 121 122
		return -EINVAL;
	}

	if (bitmap->bp[page].hijacked) /* it's hijacked, don't try to alloc */
		return 0;

	if (bitmap->bp[page].map) /* page is already allocated, just return */
		return 0;

	if (!create)
		return -ENOENT;

	/* this page has not been allocated yet */

123 124 125 126 127
	spin_unlock_irq(&bitmap->lock);
	mappage = bitmap_alloc_page(bitmap);
	spin_lock_irq(&bitmap->lock);

	if (mappage == NULL) {
128 129
		pr_debug("%s: bitmap map page allocation failed, hijacking\n",
			 bmname(bitmap));
130 131 132 133
		/* failed - set the hijacked flag so that we can use the
		 * pointer as a counter */
		if (!bitmap->bp[page].map)
			bitmap->bp[page].hijacked = 1;
134 135
	} else if (bitmap->bp[page].map ||
		   bitmap->bp[page].hijacked) {
136 137 138
		/* somebody beat us to getting the page */
		bitmap_free_page(bitmap, mappage);
		return 0;
139
	} else {
140

141
		/* no page was in place and we have one, so install it */
142

143 144 145
		bitmap->bp[page].map = mappage;
		bitmap->missing_pages--;
	}
146 147 148 149 150 151
	return 0;
}

/* if page is completely empty, put it back on the free list, or dealloc it */
/* if page was hijacked, unmark the flag so it might get alloced next time */
/* Note: lock should be held when calling this */
152
static void bitmap_checkfree(struct bitmap *bitmap, unsigned long page)
153 154 155 156 157 158 159 160 161 162 163
{
	char *ptr;

	if (bitmap->bp[page].count) /* page is still busy */
		return;

	/* page is no longer in use, it can be released */

	if (bitmap->bp[page].hijacked) { /* page was hijacked, undo this now */
		bitmap->bp[page].hijacked = 0;
		bitmap->bp[page].map = NULL;
164 165 166 167 168 169
	} else {
		/* normal case, free the page */
		ptr = bitmap->bp[page].map;
		bitmap->bp[page].map = NULL;
		bitmap->missing_pages++;
		bitmap_free_page(bitmap, ptr);
170 171 172 173 174 175 176 177 178 179 180
	}
}

/*
 * bitmap file handling - read and write the bitmap file and its superblock
 */

/*
 * basic page I/O operations
 */

181
/* IO operations when bitmap is stored near all superblocks */
182
static struct page *read_sb_page(mddev_t *mddev, loff_t offset,
183 184
				 struct page *page,
				 unsigned long index, int size)
185 186 187
{
	/* choose a good rdev and read the page from there */

188
	struct md_rdev *rdev;
189
	sector_t target;
190
	int did_alloc = 0;
191

192
	if (!page) {
193
		page = alloc_page(GFP_KERNEL);
194 195 196 197
		if (!page)
			return ERR_PTR(-ENOMEM);
		did_alloc = 1;
	}
198

199
	list_for_each_entry(rdev, &mddev->disks, same_set) {
200 201
		if (! test_bit(In_sync, &rdev->flags)
		    || test_bit(Faulty, &rdev->flags))
202 203
			continue;

J
Jonathan Brassow 已提交
204
		target = offset + index * (PAGE_SIZE/512);
205

206
		if (sync_page_io(rdev, target,
207
				 roundup(size, bdev_logical_block_size(rdev->bdev)),
J
Jonathan Brassow 已提交
208
				 page, READ, true)) {
209
			page->index = index;
210 211
			attach_page_buffers(page, NULL); /* so that free_buffer will
							  * quietly no-op */
212 213 214
			return page;
		}
	}
215 216
	if (did_alloc)
		put_page(page);
217
	return ERR_PTR(-EIO);
218 219 220

}

221
static struct md_rdev *next_active_rdev(struct md_rdev *rdev, mddev_t *mddev)
222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241
{
	/* Iterate the disks of an mddev, using rcu to protect access to the
	 * linked list, and raising the refcount of devices we return to ensure
	 * they don't disappear while in use.
	 * As devices are only added or removed when raid_disk is < 0 and
	 * nr_pending is 0 and In_sync is clear, the entries we return will
	 * still be in the same position on the list when we re-enter
	 * list_for_each_continue_rcu.
	 */
	struct list_head *pos;
	rcu_read_lock();
	if (rdev == NULL)
		/* start at the beginning */
		pos = &mddev->disks;
	else {
		/* release the previous rdev and start from there. */
		rdev_dec_pending(rdev, mddev);
		pos = &rdev->same_set;
	}
	list_for_each_continue_rcu(pos, &mddev->disks) {
242
		rdev = list_entry(pos, struct md_rdev, same_set);
243 244 245 246 247 248 249 250 251 252 253 254
		if (rdev->raid_disk >= 0 &&
		    !test_bit(Faulty, &rdev->flags)) {
			/* this is a usable devices */
			atomic_inc(&rdev->nr_pending);
			rcu_read_unlock();
			return rdev;
		}
	}
	rcu_read_unlock();
	return NULL;
}

255
static int write_sb_page(struct bitmap *bitmap, struct page *page, int wait)
256
{
257
	struct md_rdev *rdev = NULL;
258
	struct block_device *bdev;
259
	mddev_t *mddev = bitmap->mddev;
260

261
	while ((rdev = next_active_rdev(rdev, mddev)) != NULL) {
262 263
		int size = PAGE_SIZE;
		loff_t offset = mddev->bitmap_info.offset;
264 265 266

		bdev = (rdev->meta_bdev) ? rdev->meta_bdev : rdev->bdev;

267 268
		if (page->index == bitmap->file_pages-1)
			size = roundup(bitmap->last_page_size,
269
				       bdev_logical_block_size(bdev));
270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309
		/* Just make sure we aren't corrupting data or
		 * metadata
		 */
		if (mddev->external) {
			/* Bitmap could be anywhere. */
			if (rdev->sb_start + offset + (page->index
						       * (PAGE_SIZE/512))
			    > rdev->data_offset
			    &&
			    rdev->sb_start + offset
			    < (rdev->data_offset + mddev->dev_sectors
			     + (PAGE_SIZE/512)))
				goto bad_alignment;
		} else if (offset < 0) {
			/* DATA  BITMAP METADATA  */
			if (offset
			    + (long)(page->index * (PAGE_SIZE/512))
			    + size/512 > 0)
				/* bitmap runs in to metadata */
				goto bad_alignment;
			if (rdev->data_offset + mddev->dev_sectors
			    > rdev->sb_start + offset)
				/* data runs in to bitmap */
				goto bad_alignment;
		} else if (rdev->sb_start < rdev->data_offset) {
			/* METADATA BITMAP DATA */
			if (rdev->sb_start
			    + offset
			    + page->index*(PAGE_SIZE/512) + size/512
			    > rdev->data_offset)
				/* bitmap runs in to data */
				goto bad_alignment;
		} else {
			/* DATA METADATA BITMAP - no problems */
		}
		md_super_write(mddev, rdev,
			       rdev->sb_start + offset
			       + page->index * (PAGE_SIZE/512),
			       size,
			       page);
310
	}
311 312

	if (wait)
313
		md_super_wait(mddev);
314
	return 0;
315 316 317

 bad_alignment:
	return -EINVAL;
318 319
}

320
static void bitmap_file_kick(struct bitmap *bitmap);
321
/*
322
 * write out a page to a file
323
 */
324
static void write_page(struct bitmap *bitmap, struct page *page, int wait)
325
{
326
	struct buffer_head *bh;
327

328 329 330 331 332
	if (bitmap->file == NULL) {
		switch (write_sb_page(bitmap, page, wait)) {
		case -EINVAL:
			bitmap->flags |= BITMAP_WRITE_ERROR;
		}
333
	} else {
334

335
		bh = page_buffers(page);
336

337 338 339 340
		while (bh && bh->b_blocknr) {
			atomic_inc(&bitmap->pending_writes);
			set_buffer_locked(bh);
			set_buffer_mapped(bh);
J
Jens Axboe 已提交
341
			submit_bh(WRITE | REQ_SYNC, bh);
342 343
			bh = bh->b_this_page;
		}
344

345
		if (wait)
346 347
			wait_event(bitmap->write_wait,
				   atomic_read(&bitmap->pending_writes)==0);
348
	}
349 350
	if (bitmap->flags & BITMAP_WRITE_ERROR)
		bitmap_file_kick(bitmap);
351 352 353 354 355 356
}

static void end_bitmap_write(struct buffer_head *bh, int uptodate)
{
	struct bitmap *bitmap = bh->b_private;
	unsigned long flags;
357

358 359 360 361
	if (!uptodate) {
		spin_lock_irqsave(&bitmap->lock, flags);
		bitmap->flags |= BITMAP_WRITE_ERROR;
		spin_unlock_irqrestore(&bitmap->lock, flags);
362
	}
363 364 365
	if (atomic_dec_and_test(&bitmap->pending_writes))
		wake_up(&bitmap->write_wait);
}
366

367 368 369 370 371 372 373 374 375 376 377
/* copied from buffer.c */
static void
__clear_page_buffers(struct page *page)
{
	ClearPagePrivate(page);
	set_page_private(page, 0);
	page_cache_release(page);
}
static void free_buffers(struct page *page)
{
	struct buffer_head *bh = page_buffers(page);
378

379 380 381 382
	while (bh) {
		struct buffer_head *next = bh->b_this_page;
		free_buffer_head(bh);
		bh = next;
383
	}
384 385
	__clear_page_buffers(page);
	put_page(page);
386 387
}

388 389 390 391 392 393 394
/* read a page from a file.
 * We both read the page, and attach buffers to the page to record the
 * address of each block (using bmap).  These addresses will be used
 * to write the block later, completely bypassing the filesystem.
 * This usage is similar to how swap files are handled, and allows us
 * to write to a file with no concerns of memory allocation failing.
 */
395
static struct page *read_page(struct file *file, unsigned long index,
396 397
			      struct bitmap *bitmap,
			      unsigned long count)
398 399
{
	struct page *page = NULL;
J
Josef Sipek 已提交
400
	struct inode *inode = file->f_path.dentry->d_inode;
401 402
	struct buffer_head *bh;
	sector_t block;
403

404 405
	pr_debug("read bitmap file (%dB @ %llu)\n", (int)PAGE_SIZE,
		 (unsigned long long)index << PAGE_SHIFT);
406

407 408 409
	page = alloc_page(GFP_KERNEL);
	if (!page)
		page = ERR_PTR(-ENOMEM);
410 411
	if (IS_ERR(page))
		goto out;
412 413 414

	bh = alloc_page_buffers(page, 1<<inode->i_blkbits, 0);
	if (!bh) {
415
		put_page(page);
416
		page = ERR_PTR(-ENOMEM);
417 418
		goto out;
	}
419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439
	attach_page_buffers(page, bh);
	block = index << (PAGE_SHIFT - inode->i_blkbits);
	while (bh) {
		if (count == 0)
			bh->b_blocknr = 0;
		else {
			bh->b_blocknr = bmap(inode, block);
			if (bh->b_blocknr == 0) {
				/* Cannot use this file! */
				free_buffers(page);
				page = ERR_PTR(-EINVAL);
				goto out;
			}
			bh->b_bdev = inode->i_sb->s_bdev;
			if (count < (1<<inode->i_blkbits))
				count = 0;
			else
				count -= (1<<inode->i_blkbits);

			bh->b_end_io = end_bitmap_write;
			bh->b_private = bitmap;
440 441 442 443
			atomic_inc(&bitmap->pending_writes);
			set_buffer_locked(bh);
			set_buffer_mapped(bh);
			submit_bh(READ, bh);
444 445 446 447 448
		}
		block++;
		bh = bh->b_this_page;
	}
	page->index = index;
449 450 451 452 453 454 455

	wait_event(bitmap->write_wait,
		   atomic_read(&bitmap->pending_writes)==0);
	if (bitmap->flags & BITMAP_WRITE_ERROR) {
		free_buffers(page);
		page = ERR_PTR(-EIO);
	}
456 457
out:
	if (IS_ERR(page))
458
		printk(KERN_ALERT "md: bitmap read error: (%dB @ %llu): %ld\n",
459 460
			(int)PAGE_SIZE,
			(unsigned long long)index << PAGE_SHIFT,
461 462 463 464 465 466 467 468 469
			PTR_ERR(page));
	return page;
}

/*
 * bitmap file superblock operations
 */

/* update the event counter and sync the superblock to disk */
470
void bitmap_update_sb(struct bitmap *bitmap)
471 472 473 474 475
{
	bitmap_super_t *sb;
	unsigned long flags;

	if (!bitmap || !bitmap->mddev) /* no bitmap for this array */
476
		return;
477 478
	if (bitmap->mddev->bitmap_info.external)
		return;
479 480 481
	spin_lock_irqsave(&bitmap->lock, flags);
	if (!bitmap->sb_page) { /* no superblock */
		spin_unlock_irqrestore(&bitmap->lock, flags);
482
		return;
483 484
	}
	spin_unlock_irqrestore(&bitmap->lock, flags);
485
	sb = kmap_atomic(bitmap->sb_page, KM_USER0);
486
	sb->events = cpu_to_le64(bitmap->mddev->events);
487
	if (bitmap->mddev->events < bitmap->events_cleared)
488 489
		/* rocking back to read-only */
		bitmap->events_cleared = bitmap->mddev->events;
490 491
	sb->events_cleared = cpu_to_le64(bitmap->events_cleared);
	sb->state = cpu_to_le32(bitmap->flags);
492 493 494
	/* Just in case these have been changed via sysfs: */
	sb->daemon_sleep = cpu_to_le32(bitmap->mddev->bitmap_info.daemon_sleep/HZ);
	sb->write_behind = cpu_to_le32(bitmap->mddev->bitmap_info.max_write_behind);
495
	kunmap_atomic(sb, KM_USER0);
496
	write_page(bitmap, bitmap->sb_page, 1);
497 498 499 500 501 502 503 504 505
}

/* print out the bitmap file superblock */
void bitmap_print_sb(struct bitmap *bitmap)
{
	bitmap_super_t *sb;

	if (!bitmap || !bitmap->sb_page)
		return;
506
	sb = kmap_atomic(bitmap->sb_page, KM_USER0);
507
	printk(KERN_DEBUG "%s: bitmap file superblock:\n", bmname(bitmap));
508 509 510
	printk(KERN_DEBUG "         magic: %08x\n", le32_to_cpu(sb->magic));
	printk(KERN_DEBUG "       version: %d\n", le32_to_cpu(sb->version));
	printk(KERN_DEBUG "          uuid: %08x.%08x.%08x.%08x\n",
511 512 513 514
					*(__u32 *)(sb->uuid+0),
					*(__u32 *)(sb->uuid+4),
					*(__u32 *)(sb->uuid+8),
					*(__u32 *)(sb->uuid+12));
515
	printk(KERN_DEBUG "        events: %llu\n",
516
			(unsigned long long) le64_to_cpu(sb->events));
517
	printk(KERN_DEBUG "events cleared: %llu\n",
518
			(unsigned long long) le64_to_cpu(sb->events_cleared));
519 520 521 522 523
	printk(KERN_DEBUG "         state: %08x\n", le32_to_cpu(sb->state));
	printk(KERN_DEBUG "     chunksize: %d B\n", le32_to_cpu(sb->chunksize));
	printk(KERN_DEBUG "  daemon sleep: %ds\n", le32_to_cpu(sb->daemon_sleep));
	printk(KERN_DEBUG "     sync size: %llu KB\n",
			(unsigned long long)le64_to_cpu(sb->sync_size)/2);
524
	printk(KERN_DEBUG "max write behind: %d\n", le32_to_cpu(sb->write_behind));
525
	kunmap_atomic(sb, KM_USER0);
526 527
}

528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603
/*
 * bitmap_new_disk_sb
 * @bitmap
 *
 * This function is somewhat the reverse of bitmap_read_sb.  bitmap_read_sb
 * reads and verifies the on-disk bitmap superblock and populates bitmap_info.
 * This function verifies 'bitmap_info' and populates the on-disk bitmap
 * structure, which is to be written to disk.
 *
 * Returns: 0 on success, -Exxx on error
 */
static int bitmap_new_disk_sb(struct bitmap *bitmap)
{
	bitmap_super_t *sb;
	unsigned long chunksize, daemon_sleep, write_behind;
	int err = -EINVAL;

	bitmap->sb_page = alloc_page(GFP_KERNEL);
	if (IS_ERR(bitmap->sb_page)) {
		err = PTR_ERR(bitmap->sb_page);
		bitmap->sb_page = NULL;
		return err;
	}
	bitmap->sb_page->index = 0;

	sb = kmap_atomic(bitmap->sb_page, KM_USER0);

	sb->magic = cpu_to_le32(BITMAP_MAGIC);
	sb->version = cpu_to_le32(BITMAP_MAJOR_HI);

	chunksize = bitmap->mddev->bitmap_info.chunksize;
	BUG_ON(!chunksize);
	if (!is_power_of_2(chunksize)) {
		kunmap_atomic(sb, KM_USER0);
		printk(KERN_ERR "bitmap chunksize not a power of 2\n");
		return -EINVAL;
	}
	sb->chunksize = cpu_to_le32(chunksize);

	daemon_sleep = bitmap->mddev->bitmap_info.daemon_sleep;
	if (!daemon_sleep ||
	    (daemon_sleep < 1) || (daemon_sleep > MAX_SCHEDULE_TIMEOUT)) {
		printk(KERN_INFO "Choosing daemon_sleep default (5 sec)\n");
		daemon_sleep = 5 * HZ;
	}
	sb->daemon_sleep = cpu_to_le32(daemon_sleep);
	bitmap->mddev->bitmap_info.daemon_sleep = daemon_sleep;

	/*
	 * FIXME: write_behind for RAID1.  If not specified, what
	 * is a good choice?  We choose COUNTER_MAX / 2 arbitrarily.
	 */
	write_behind = bitmap->mddev->bitmap_info.max_write_behind;
	if (write_behind > COUNTER_MAX)
		write_behind = COUNTER_MAX / 2;
	sb->write_behind = cpu_to_le32(write_behind);
	bitmap->mddev->bitmap_info.max_write_behind = write_behind;

	/* keep the array size field of the bitmap superblock up to date */
	sb->sync_size = cpu_to_le64(bitmap->mddev->resync_max_sectors);

	memcpy(sb->uuid, bitmap->mddev->uuid, 16);

	bitmap->flags |= BITMAP_STALE;
	sb->state |= cpu_to_le32(BITMAP_STALE);
	bitmap->events_cleared = bitmap->mddev->events;
	sb->events_cleared = cpu_to_le64(bitmap->mddev->events);

	bitmap->flags |= BITMAP_HOSTENDIAN;
	sb->version = cpu_to_le32(BITMAP_MAJOR_HOSTENDIAN);

	kunmap_atomic(sb, KM_USER0);

	return 0;
}

604 605 606 607 608
/* read the superblock from the bitmap file and initialize some bitmap fields */
static int bitmap_read_sb(struct bitmap *bitmap)
{
	char *reason = NULL;
	bitmap_super_t *sb;
609
	unsigned long chunksize, daemon_sleep, write_behind;
610 611 612 613
	unsigned long long events;
	int err = -EINVAL;

	/* page 0 is the superblock, read it... */
614 615 616 617 618 619
	if (bitmap->file) {
		loff_t isize = i_size_read(bitmap->file->f_mapping->host);
		int bytes = isize > PAGE_SIZE ? PAGE_SIZE : isize;

		bitmap->sb_page = read_page(bitmap->file, 0, bitmap, bytes);
	} else {
620 621
		bitmap->sb_page = read_sb_page(bitmap->mddev,
					       bitmap->mddev->bitmap_info.offset,
622 623
					       NULL,
					       0, sizeof(bitmap_super_t));
624
	}
625 626 627 628 629 630
	if (IS_ERR(bitmap->sb_page)) {
		err = PTR_ERR(bitmap->sb_page);
		bitmap->sb_page = NULL;
		return err;
	}

631
	sb = kmap_atomic(bitmap->sb_page, KM_USER0);
632 633

	chunksize = le32_to_cpu(sb->chunksize);
634
	daemon_sleep = le32_to_cpu(sb->daemon_sleep) * HZ;
635
	write_behind = le32_to_cpu(sb->write_behind);
636 637 638 639

	/* verify that the bitmap-specific fields are valid */
	if (sb->magic != cpu_to_le32(BITMAP_MAGIC))
		reason = "bad magic";
640 641
	else if (le32_to_cpu(sb->version) < BITMAP_MAJOR_LO ||
		 le32_to_cpu(sb->version) > BITMAP_MAJOR_HI)
642
		reason = "unrecognized superblock version";
643
	else if (chunksize < 512)
644
		reason = "bitmap chunksize too small";
J
Jonathan Brassow 已提交
645
	else if (!is_power_of_2(chunksize))
646
		reason = "bitmap chunksize not a power of 2";
647
	else if (daemon_sleep < 1 || daemon_sleep > MAX_SCHEDULE_TIMEOUT)
648
		reason = "daemon sleep period out of range";
649 650
	else if (write_behind > COUNTER_MAX)
		reason = "write-behind limit out of range (0 - 16383)";
651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676
	if (reason) {
		printk(KERN_INFO "%s: invalid bitmap file superblock: %s\n",
			bmname(bitmap), reason);
		goto out;
	}

	/* keep the array size field of the bitmap superblock up to date */
	sb->sync_size = cpu_to_le64(bitmap->mddev->resync_max_sectors);

	if (!bitmap->mddev->persistent)
		goto success;

	/*
	 * if we have a persistent array superblock, compare the
	 * bitmap's UUID and event counter to the mddev's
	 */
	if (memcmp(sb->uuid, bitmap->mddev->uuid, 16)) {
		printk(KERN_INFO "%s: bitmap superblock UUID mismatch\n",
			bmname(bitmap));
		goto out;
	}
	events = le64_to_cpu(sb->events);
	if (events < bitmap->mddev->events) {
		printk(KERN_INFO "%s: bitmap file is out of date (%llu < %llu) "
			"-- forcing full recovery\n", bmname(bitmap), events,
			(unsigned long long) bitmap->mddev->events);
677
		sb->state |= cpu_to_le32(BITMAP_STALE);
678 679 680
	}
success:
	/* assign fields using values from superblock */
681 682 683
	bitmap->mddev->bitmap_info.chunksize = chunksize;
	bitmap->mddev->bitmap_info.daemon_sleep = daemon_sleep;
	bitmap->mddev->bitmap_info.max_write_behind = write_behind;
684
	bitmap->flags |= le32_to_cpu(sb->state);
685 686
	if (le32_to_cpu(sb->version) == BITMAP_MAJOR_HOSTENDIAN)
		bitmap->flags |= BITMAP_HOSTENDIAN;
687
	bitmap->events_cleared = le64_to_cpu(sb->events_cleared);
688
	if (bitmap->flags & BITMAP_STALE)
689
		bitmap->events_cleared = bitmap->mddev->events;
690 691
	err = 0;
out:
692
	kunmap_atomic(sb, KM_USER0);
693 694 695 696 697 698 699 700 701 702
	if (err)
		bitmap_print_sb(bitmap);
	return err;
}

enum bitmap_mask_op {
	MASK_SET,
	MASK_UNSET
};

703 704 705
/* record the state of the bitmap in the superblock.  Return the old value */
static int bitmap_mask_state(struct bitmap *bitmap, enum bitmap_state bits,
			     enum bitmap_mask_op op)
706 707 708
{
	bitmap_super_t *sb;
	unsigned long flags;
709
	int old;
710 711

	spin_lock_irqsave(&bitmap->lock, flags);
712
	if (!bitmap->sb_page) { /* can't set the state */
713
		spin_unlock_irqrestore(&bitmap->lock, flags);
714
		return 0;
715 716
	}
	spin_unlock_irqrestore(&bitmap->lock, flags);
717
	sb = kmap_atomic(bitmap->sb_page, KM_USER0);
718
	old = le32_to_cpu(sb->state) & bits;
719
	switch (op) {
720 721
	case MASK_SET:
		sb->state |= cpu_to_le32(bits);
722
		bitmap->flags |= bits;
723 724 725
		break;
	case MASK_UNSET:
		sb->state &= cpu_to_le32(~bits);
726
		bitmap->flags &= ~bits;
727 728 729
		break;
	default:
		BUG();
730
	}
731
	kunmap_atomic(sb, KM_USER0);
732
	return old;
733 734 735 736 737 738
}

/*
 * general bitmap file operations
 */

739 740 741 742 743 744
/*
 * on-disk bitmap:
 *
 * Use one bit per "chunk" (block set). We do the disk I/O on the bitmap
 * file a page at a time. There's a superblock at the start of the file.
 */
745
/* calculate the index of the page that contains this bit */
746
static inline unsigned long file_page_index(struct bitmap *bitmap, unsigned long chunk)
747
{
748 749 750
	if (!bitmap->mddev->bitmap_info.external)
		chunk += sizeof(bitmap_super_t) << 3;
	return chunk >> PAGE_BIT_SHIFT;
751 752 753
}

/* calculate the (bit) offset of this bit within a page */
754
static inline unsigned long file_page_offset(struct bitmap *bitmap, unsigned long chunk)
755
{
756 757 758
	if (!bitmap->mddev->bitmap_info.external)
		chunk += sizeof(bitmap_super_t) << 3;
	return chunk & (PAGE_BITS - 1);
759 760 761 762 763 764 765 766 767 768
}

/*
 * return a pointer to the page in the filemap that contains the given bit
 *
 * this lookup is complicated by the fact that the bitmap sb might be exactly
 * 1 page (e.g., x86) or less than 1 page -- so the bitmap might start on page
 * 0 or page 1
 */
static inline struct page *filemap_get_page(struct bitmap *bitmap,
769
					    unsigned long chunk)
770
{
771 772
	if (file_page_index(bitmap, chunk) >= bitmap->file_pages)
		return NULL;
773 774
	return bitmap->filemap[file_page_index(bitmap, chunk)
			       - file_page_index(bitmap, 0)];
775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795
}

static void bitmap_file_unmap(struct bitmap *bitmap)
{
	struct page **map, *sb_page;
	unsigned long *attr;
	int pages;
	unsigned long flags;

	spin_lock_irqsave(&bitmap->lock, flags);
	map = bitmap->filemap;
	bitmap->filemap = NULL;
	attr = bitmap->filemap_attr;
	bitmap->filemap_attr = NULL;
	pages = bitmap->file_pages;
	bitmap->file_pages = 0;
	sb_page = bitmap->sb_page;
	bitmap->sb_page = NULL;
	spin_unlock_irqrestore(&bitmap->lock, flags);

	while (pages--)
796
		if (map[pages] != sb_page) /* 0 is sb_page, release it below */
797
			free_buffers(map[pages]);
798 799 800
	kfree(map);
	kfree(attr);

801 802
	if (sb_page)
		free_buffers(sb_page);
803 804 805 806 807 808 809 810 811 812 813 814
}

static void bitmap_file_put(struct bitmap *bitmap)
{
	struct file *file;
	unsigned long flags;

	spin_lock_irqsave(&bitmap->lock, flags);
	file = bitmap->file;
	bitmap->file = NULL;
	spin_unlock_irqrestore(&bitmap->lock, flags);

815 816 817
	if (file)
		wait_event(bitmap->write_wait,
			   atomic_read(&bitmap->pending_writes)==0);
818 819
	bitmap_file_unmap(bitmap);

820
	if (file) {
J
Josef Sipek 已提交
821
		struct inode *inode = file->f_path.dentry->d_inode;
822
		invalidate_mapping_pages(inode->i_mapping, 0, -1);
823
		fput(file);
824
	}
825 826 827 828 829 830 831 832 833 834 835
}

/*
 * bitmap_file_kick - if an error occurs while manipulating the bitmap file
 * then it is no longer reliable, so we stop using it and we mark the file
 * as failed in the superblock
 */
static void bitmap_file_kick(struct bitmap *bitmap)
{
	char *path, *ptr = NULL;

836 837
	if (bitmap_mask_state(bitmap, BITMAP_STALE, MASK_SET) == 0) {
		bitmap_update_sb(bitmap);
838

839 840 841
		if (bitmap->file) {
			path = kmalloc(PAGE_SIZE, GFP_KERNEL);
			if (path)
C
Christoph Hellwig 已提交
842 843 844
				ptr = d_path(&bitmap->file->f_path, path,
					     PAGE_SIZE);

845 846
			printk(KERN_ALERT
			      "%s: kicking failed bitmap file %s from array!\n",
C
Christoph Hellwig 已提交
847
			      bmname(bitmap), IS_ERR(ptr) ? "" : ptr);
848

849 850 851 852 853
			kfree(path);
		} else
			printk(KERN_ALERT
			       "%s: disabling internal bitmap due to errors\n",
			       bmname(bitmap));
854
	}
855 856 857 858 859 860 861

	bitmap_file_put(bitmap);

	return;
}

enum bitmap_page_attr {
862
	BITMAP_PAGE_DIRTY = 0,     /* there are set bits that need to be synced */
863 864
	BITMAP_PAGE_PENDING = 1,   /* there are bits that are being cleaned.
				    * i.e. counter is 1 or 2. */
865
	BITMAP_PAGE_NEEDWRITE = 2, /* there are cleared bits that need to be synced */
866 867 868 869 870
};

static inline void set_page_attr(struct bitmap *bitmap, struct page *page,
				enum bitmap_page_attr attr)
{
871
	__set_bit((page->index<<2) + attr, bitmap->filemap_attr);
872 873 874 875 876
}

static inline void clear_page_attr(struct bitmap *bitmap, struct page *page,
				enum bitmap_page_attr attr)
{
877
	__clear_bit((page->index<<2) + attr, bitmap->filemap_attr);
878 879
}

880 881
static inline unsigned long test_page_attr(struct bitmap *bitmap, struct page *page,
					   enum bitmap_page_attr attr)
882
{
883
	return test_bit((page->index<<2) + attr, bitmap->filemap_attr);
884 885 886 887 888 889 890 891 892 893 894 895
}

/*
 * bitmap_file_set_bit -- called before performing a write to the md device
 * to set (and eventually sync) a particular bit in the bitmap file
 *
 * we set the bit immediately, then we record the page number so that
 * when an unplug occurs, we can flush the dirty pages out to disk
 */
static void bitmap_file_set_bit(struct bitmap *bitmap, sector_t block)
{
	unsigned long bit;
896
	struct page *page;
897 898 899
	void *kaddr;
	unsigned long chunk = block >> CHUNK_BLOCK_SHIFT(bitmap);

900 901
	if (!bitmap->filemap)
		return;
902

903 904 905 906
	page = filemap_get_page(bitmap, chunk);
	if (!page)
		return;
	bit = file_page_offset(bitmap, chunk);
907

908 909 910 911 912 913 914
	/* set the bit */
	kaddr = kmap_atomic(page, KM_USER0);
	if (bitmap->flags & BITMAP_HOSTENDIAN)
		set_bit(bit, kaddr);
	else
		__set_bit_le(bit, kaddr);
	kunmap_atomic(kaddr, KM_USER0);
915
	pr_debug("set file bit %lu page %lu\n", bit, page->index);
916 917 918 919 920 921 922
	/* record page number so it gets flushed to disk when unplug occurs */
	set_page_attr(bitmap, page, BITMAP_PAGE_DIRTY);
}

/* this gets called when the md device is ready to unplug its underlying
 * (slave) device queues -- before we let any writes go down, we need to
 * sync the dirty pages of the bitmap file to disk */
923
void bitmap_unplug(struct bitmap *bitmap)
924
{
925 926
	unsigned long i, flags;
	int dirty, need_write;
927 928 929 930
	struct page *page;
	int wait = 0;

	if (!bitmap)
931
		return;
932 933 934 935 936

	/* look at each page to see if there are any set bits that need to be
	 * flushed out to disk */
	for (i = 0; i < bitmap->file_pages; i++) {
		spin_lock_irqsave(&bitmap->lock, flags);
937
		if (!bitmap->filemap) {
938
			spin_unlock_irqrestore(&bitmap->lock, flags);
939
			return;
940 941
		}
		page = bitmap->filemap[i];
942 943
		dirty = test_page_attr(bitmap, page, BITMAP_PAGE_DIRTY);
		need_write = test_page_attr(bitmap, page, BITMAP_PAGE_NEEDWRITE);
944 945
		clear_page_attr(bitmap, page, BITMAP_PAGE_DIRTY);
		clear_page_attr(bitmap, page, BITMAP_PAGE_NEEDWRITE);
946
		if (dirty)
947 948 949
			wait = 1;
		spin_unlock_irqrestore(&bitmap->lock, flags);

950
		if (dirty || need_write)
951
			write_page(bitmap, page, 0);
952 953
	}
	if (wait) { /* if any writes were performed, we need to wait on them */
954
		if (bitmap->file)
955 956
			wait_event(bitmap->write_wait,
				   atomic_read(&bitmap->pending_writes)==0);
957
		else
958
			md_super_wait(bitmap->mddev);
959
	}
960 961
	if (bitmap->flags & BITMAP_WRITE_ERROR)
		bitmap_file_kick(bitmap);
962
}
963
EXPORT_SYMBOL(bitmap_unplug);
964

965
static void bitmap_set_memory_bits(struct bitmap *bitmap, sector_t offset, int needed);
966 967 968 969 970 971 972
/* * bitmap_init_from_disk -- called at bitmap_create time to initialize
 * the in-memory bitmap from the on-disk bitmap -- also, sets up the
 * memory mapping of the bitmap file
 * Special cases:
 *   if there's no bitmap file, or if the bitmap file had been
 *   previously kicked from the array, we mark all the bits as
 *   1's in order to cause a full resync.
973 974 975
 *
 * We ignore all bits for sectors that end earlier than 'start'.
 * This is used when reading an out-of-date bitmap...
976
 */
977
static int bitmap_init_from_disk(struct bitmap *bitmap, sector_t start)
978 979 980 981 982
{
	unsigned long i, chunks, index, oldindex, bit;
	struct page *page = NULL, *oldpage = NULL;
	unsigned long num_pages, bit_cnt = 0;
	struct file *file;
983
	unsigned long bytes, offset;
984 985
	int outofdate;
	int ret = -ENOSPC;
986
	void *paddr;
987 988 989 990

	chunks = bitmap->chunks;
	file = bitmap->file;

991
	BUG_ON(!file && !bitmap->mddev->bitmap_info.offset);
992

993
#ifdef INJECT_FAULTS_3
994 995 996 997 998 999 1000 1001
	outofdate = 1;
#else
	outofdate = bitmap->flags & BITMAP_STALE;
#endif
	if (outofdate)
		printk(KERN_INFO "%s: bitmap file is out of date, doing full "
			"recovery\n", bmname(bitmap));

1002
	bytes = DIV_ROUND_UP(bitmap->chunks, 8);
1003 1004
	if (!bitmap->mddev->bitmap_info.external)
		bytes += sizeof(bitmap_super_t);
1005

1006
	num_pages = DIV_ROUND_UP(bytes, PAGE_SIZE);
1007

1008
	if (file && i_size_read(file->f_mapping->host) < bytes) {
1009 1010 1011
		printk(KERN_INFO "%s: bitmap file too short %lu < %lu\n",
			bmname(bitmap),
			(unsigned long) i_size_read(file->f_mapping->host),
1012
			bytes);
1013
		goto err;
1014
	}
1015 1016 1017

	ret = -ENOMEM;

1018
	bitmap->filemap = kmalloc(sizeof(struct page *) * num_pages, GFP_KERNEL);
1019
	if (!bitmap->filemap)
1020
		goto err;
1021

1022 1023
	/* We need 4 bits per page, rounded up to a multiple of sizeof(unsigned long) */
	bitmap->filemap_attr = kzalloc(
1024
		roundup(DIV_ROUND_UP(num_pages*4, 8), sizeof(unsigned long)),
1025
		GFP_KERNEL);
1026
	if (!bitmap->filemap_attr)
1027
		goto err;
1028 1029 1030 1031

	oldindex = ~0L;

	for (i = 0; i < chunks; i++) {
1032
		int b;
1033 1034
		index = file_page_index(bitmap, i);
		bit = file_page_offset(bitmap, i);
1035
		if (index != oldindex) { /* this is a new page, read it in */
1036
			int count;
1037
			/* unmap the old page, we're done with it */
1038
			if (index == num_pages-1)
1039
				count = bytes - index * PAGE_SIZE;
1040 1041
			else
				count = PAGE_SIZE;
1042
			if (index == 0 && bitmap->sb_page) {
1043 1044 1045 1046 1047 1048 1049
				/*
				 * if we're here then the superblock page
				 * contains some bits (PAGE_SIZE != sizeof sb)
				 * we've already read it in, so just use it
				 */
				page = bitmap->sb_page;
				offset = sizeof(bitmap_super_t);
N
NeilBrown 已提交
1050
				if (!file)
1051 1052 1053 1054 1055
					page = read_sb_page(
						bitmap->mddev,
						bitmap->mddev->bitmap_info.offset,
						page,
						index, count);
1056
			} else if (file) {
1057
				page = read_page(file, index, bitmap, count);
1058 1059
				offset = 0;
			} else {
1060 1061
				page = read_sb_page(bitmap->mddev,
						    bitmap->mddev->bitmap_info.offset,
1062 1063
						    NULL,
						    index, count);
1064 1065
				offset = 0;
			}
1066 1067
			if (IS_ERR(page)) { /* read error */
				ret = PTR_ERR(page);
1068
				goto err;
1069 1070
			}

1071 1072 1073
			oldindex = index;
			oldpage = page;

1074 1075 1076
			bitmap->filemap[bitmap->file_pages++] = page;
			bitmap->last_page_size = count;

1077 1078 1079
			if (outofdate) {
				/*
				 * if bitmap is out of date, dirty the
1080
				 * whole page and write it out
1081
				 */
1082 1083
				paddr = kmap_atomic(page, KM_USER0);
				memset(paddr + offset, 0xff,
1084
				       PAGE_SIZE - offset);
1085
				kunmap_atomic(paddr, KM_USER0);
1086 1087 1088
				write_page(bitmap, page, 1);

				ret = -EIO;
1089
				if (bitmap->flags & BITMAP_WRITE_ERROR)
1090
					goto err;
1091 1092
			}
		}
1093
		paddr = kmap_atomic(page, KM_USER0);
1094
		if (bitmap->flags & BITMAP_HOSTENDIAN)
1095
			b = test_bit(bit, paddr);
1096
		else
A
Akinobu Mita 已提交
1097
			b = test_bit_le(bit, paddr);
1098
		kunmap_atomic(paddr, KM_USER0);
1099
		if (b) {
1100
			/* if the disk bit is set, set the memory bit */
1101 1102 1103 1104 1105
			int needed = ((sector_t)(i+1) << (CHUNK_BLOCK_SHIFT(bitmap))
				      >= start);
			bitmap_set_memory_bits(bitmap,
					       (sector_t)i << CHUNK_BLOCK_SHIFT(bitmap),
					       needed);
1106 1107 1108 1109
			bit_cnt++;
		}
	}

1110
	/* everything went OK */
1111 1112 1113 1114 1115 1116 1117 1118 1119
	ret = 0;
	bitmap_mask_state(bitmap, BITMAP_STALE, MASK_UNSET);

	if (bit_cnt) { /* Kick recovery if any bits were set */
		set_bit(MD_RECOVERY_NEEDED, &bitmap->mddev->recovery);
		md_wakeup_thread(bitmap->mddev->thread);
	}

	printk(KERN_INFO "%s: bitmap initialized from disk: "
1120 1121
	       "read %lu/%lu pages, set %lu of %lu bits\n",
	       bmname(bitmap), bitmap->file_pages, num_pages, bit_cnt, chunks);
1122 1123

	return 0;
1124

1125 1126 1127
 err:
	printk(KERN_INFO "%s: bitmap initialisation failed: %d\n",
	       bmname(bitmap), ret);
1128 1129 1130
	return ret;
}

1131 1132 1133 1134 1135
void bitmap_write_all(struct bitmap *bitmap)
{
	/* We don't actually write all bitmap blocks here,
	 * just flag them as needing to be written
	 */
1136
	int i;
1137

1138
	for (i = 0; i < bitmap->file_pages; i++)
1139 1140
		set_page_attr(bitmap, bitmap->filemap[i],
			      BITMAP_PAGE_NEEDWRITE);
1141
	bitmap->allclean = 0;
1142 1143
}

1144 1145 1146 1147 1148 1149 1150 1151
static void bitmap_count_page(struct bitmap *bitmap, sector_t offset, int inc)
{
	sector_t chunk = offset >> CHUNK_BLOCK_SHIFT(bitmap);
	unsigned long page = chunk >> PAGE_COUNTER_SHIFT;
	bitmap->bp[page].count += inc;
	bitmap_checkfree(bitmap, page);
}
static bitmap_counter_t *bitmap_get_counter(struct bitmap *bitmap,
N
NeilBrown 已提交
1152
					    sector_t offset, sector_t *blocks,
1153 1154 1155 1156 1157 1158 1159
					    int create);

/*
 * bitmap daemon -- periodically wakes up to clean bits and flush pages
 *			out to disk
 */

1160
void bitmap_daemon_work(mddev_t *mddev)
1161
{
1162
	struct bitmap *bitmap;
1163
	unsigned long j;
1164 1165
	unsigned long flags;
	struct page *page = NULL, *lastpage = NULL;
N
NeilBrown 已提交
1166
	sector_t blocks;
1167
	void *paddr;
1168

1169 1170 1171
	/* Use a mutex to guard daemon_work against
	 * bitmap_destroy.
	 */
1172
	mutex_lock(&mddev->bitmap_info.mutex);
1173 1174
	bitmap = mddev->bitmap;
	if (bitmap == NULL) {
1175
		mutex_unlock(&mddev->bitmap_info.mutex);
1176
		return;
1177
	}
1178
	if (time_before(jiffies, bitmap->daemon_lastrun
1179
			+ bitmap->mddev->bitmap_info.daemon_sleep))
1180 1181
		goto done;

1182
	bitmap->daemon_lastrun = jiffies;
1183 1184
	if (bitmap->allclean) {
		bitmap->mddev->thread->timeout = MAX_SCHEDULE_TIMEOUT;
1185
		goto done;
1186 1187
	}
	bitmap->allclean = 1;
1188

1189
	spin_lock_irqsave(&bitmap->lock, flags);
1190 1191
	for (j = 0; j < bitmap->chunks; j++) {
		bitmap_counter_t *bmc;
1192 1193 1194 1195 1196
		if (!bitmap->filemap)
			/* error or shutdown */
			break;

		page = filemap_get_page(bitmap, j);
1197 1198

		if (page != lastpage) {
1199
			/* skip this page unless it's marked as needing cleaning */
1200
			if (!test_page_attr(bitmap, page, BITMAP_PAGE_PENDING)) {
1201 1202
				int need_write = test_page_attr(bitmap, page,
								BITMAP_PAGE_NEEDWRITE);
1203
				if (need_write)
1204
					clear_page_attr(bitmap, page, BITMAP_PAGE_NEEDWRITE);
1205

1206
				spin_unlock_irqrestore(&bitmap->lock, flags);
1207
				if (need_write)
1208
					write_page(bitmap, page, 0);
1209 1210
				spin_lock_irqsave(&bitmap->lock, flags);
				j |= (PAGE_BITS - 1);
1211 1212 1213
				continue;
			}

1214 1215
			/* grab the new page, sync and release the old */
			if (lastpage != NULL) {
1216 1217 1218 1219
				if (test_page_attr(bitmap, lastpage,
						   BITMAP_PAGE_NEEDWRITE)) {
					clear_page_attr(bitmap, lastpage,
							BITMAP_PAGE_NEEDWRITE);
1220
					spin_unlock_irqrestore(&bitmap->lock, flags);
1221
					write_page(bitmap, lastpage, 0);
1222
				} else {
1223 1224 1225
					set_page_attr(bitmap, lastpage,
						      BITMAP_PAGE_NEEDWRITE);
					bitmap->allclean = 0;
1226 1227 1228 1229 1230
					spin_unlock_irqrestore(&bitmap->lock, flags);
				}
			} else
				spin_unlock_irqrestore(&bitmap->lock, flags);
			lastpage = page;
1231 1232 1233 1234

			/* We are possibly going to clear some bits, so make
			 * sure that events_cleared is up-to-date.
			 */
1235 1236
			if (bitmap->need_sync &&
			    bitmap->mddev->bitmap_info.external == 0) {
1237 1238 1239 1240 1241 1242 1243 1244
				bitmap_super_t *sb;
				bitmap->need_sync = 0;
				sb = kmap_atomic(bitmap->sb_page, KM_USER0);
				sb->events_cleared =
					cpu_to_le64(bitmap->events_cleared);
				kunmap_atomic(sb, KM_USER0);
				write_page(bitmap, bitmap->sb_page, 1);
			}
1245
			spin_lock_irqsave(&bitmap->lock, flags);
1246
			if (!bitmap->need_sync)
1247
				clear_page_attr(bitmap, page, BITMAP_PAGE_PENDING);
1248 1249
			else
				bitmap->allclean = 0;
1250
		}
1251 1252 1253
		bmc = bitmap_get_counter(bitmap,
					 (sector_t)j << CHUNK_BLOCK_SHIFT(bitmap),
					 &blocks, 0);
1254 1255 1256 1257
		if (!bmc)
			j |= PAGE_COUNTER_MASK;
		else if (*bmc) {
			if (*bmc == 1 && !bitmap->need_sync) {
1258 1259
				/* we can clear the bit */
				*bmc = 0;
1260 1261
				bitmap_count_page(bitmap,
						  (sector_t)j << CHUNK_BLOCK_SHIFT(bitmap),
1262 1263 1264
						  -1);

				/* clear the bit */
1265 1266 1267 1268 1269 1270
				paddr = kmap_atomic(page, KM_USER0);
				if (bitmap->flags & BITMAP_HOSTENDIAN)
					clear_bit(file_page_offset(bitmap, j),
						  paddr);
				else
					__clear_bit_le(
1271 1272 1273
						file_page_offset(bitmap,
								 j),
						paddr);
1274
				kunmap_atomic(paddr, KM_USER0);
1275 1276 1277
			} else if (*bmc <= 2) {
				*bmc = 1; /* maybe clear the bit next time */
				set_page_attr(bitmap, page, BITMAP_PAGE_PENDING);
1278
				bitmap->allclean = 0;
1279
			}
1280
		}
1281
	}
1282
	spin_unlock_irqrestore(&bitmap->lock, flags);
1283 1284

	/* now sync the final page */
1285
	if (lastpage != NULL) {
1286
		spin_lock_irqsave(&bitmap->lock, flags);
1287
		if (test_page_attr(bitmap, lastpage, BITMAP_PAGE_NEEDWRITE)) {
1288 1289
			clear_page_attr(bitmap, lastpage, BITMAP_PAGE_NEEDWRITE);
			spin_unlock_irqrestore(&bitmap->lock, flags);
1290
			write_page(bitmap, lastpage, 0);
1291 1292
		} else {
			set_page_attr(bitmap, lastpage, BITMAP_PAGE_NEEDWRITE);
1293
			bitmap->allclean = 0;
1294 1295 1296 1297
			spin_unlock_irqrestore(&bitmap->lock, flags);
		}
	}

1298
 done:
1299
	if (bitmap->allclean == 0)
1300
		bitmap->mddev->thread->timeout =
1301
			bitmap->mddev->bitmap_info.daemon_sleep;
1302
	mutex_unlock(&mddev->bitmap_info.mutex);
1303 1304 1305
}

static bitmap_counter_t *bitmap_get_counter(struct bitmap *bitmap,
N
NeilBrown 已提交
1306
					    sector_t offset, sector_t *blocks,
1307
					    int create)
1308 1309
__releases(bitmap->lock)
__acquires(bitmap->lock)
1310 1311 1312 1313 1314 1315 1316 1317 1318
{
	/* If 'create', we might release the lock and reclaim it.
	 * The lock must have been taken with interrupts enabled.
	 * If !create, we don't release the lock.
	 */
	sector_t chunk = offset >> CHUNK_BLOCK_SHIFT(bitmap);
	unsigned long page = chunk >> PAGE_COUNTER_SHIFT;
	unsigned long pageoff = (chunk & PAGE_COUNTER_MASK) << COUNTER_BYTE_SHIFT;
	sector_t csize;
1319
	int err;
1320

1321 1322 1323 1324 1325 1326 1327
	err = bitmap_checkpage(bitmap, page, create);

	if (bitmap->bp[page].hijacked ||
	    bitmap->bp[page].map == NULL)
		csize = ((sector_t)1) << (CHUNK_BLOCK_SHIFT(bitmap) +
					  PAGE_COUNTER_SHIFT - 1);
	else
1328
		csize = ((sector_t)1) << (CHUNK_BLOCK_SHIFT(bitmap));
1329 1330 1331
	*blocks = csize - (offset & (csize - 1));

	if (err < 0)
1332
		return NULL;
1333

1334 1335 1336 1337 1338 1339 1340 1341
	/* now locked ... */

	if (bitmap->bp[page].hijacked) { /* hijacked pointer */
		/* should we use the first or second counter field
		 * of the hijacked pointer? */
		int hi = (pageoff > PAGE_COUNTER_MASK);
		return  &((bitmap_counter_t *)
			  &bitmap->bp[page].map)[hi];
1342
	} else /* page is allocated */
1343 1344 1345 1346
		return (bitmap_counter_t *)
			&(bitmap->bp[page].map[pageoff]);
}

1347
int bitmap_startwrite(struct bitmap *bitmap, sector_t offset, unsigned long sectors, int behind)
1348
{
1349 1350
	if (!bitmap)
		return 0;
1351 1352

	if (behind) {
1353
		int bw;
1354
		atomic_inc(&bitmap->behind_writes);
1355 1356 1357 1358
		bw = atomic_read(&bitmap->behind_writes);
		if (bw > bitmap->behind_writes_used)
			bitmap->behind_writes_used = bw;

1359 1360
		pr_debug("inc write-behind count %d/%lu\n",
			 bw, bitmap->mddev->bitmap_info.max_write_behind);
1361 1362
	}

1363
	while (sectors) {
N
NeilBrown 已提交
1364
		sector_t blocks;
1365 1366 1367 1368 1369 1370 1371 1372 1373
		bitmap_counter_t *bmc;

		spin_lock_irq(&bitmap->lock);
		bmc = bitmap_get_counter(bitmap, offset, &blocks, 1);
		if (!bmc) {
			spin_unlock_irq(&bitmap->lock);
			return 0;
		}

1374
		if (unlikely(COUNTER(*bmc) == COUNTER_MAX)) {
1375 1376 1377 1378 1379 1380 1381 1382
			DEFINE_WAIT(__wait);
			/* note that it is safe to do the prepare_to_wait
			 * after the test as long as we do it before dropping
			 * the spinlock.
			 */
			prepare_to_wait(&bitmap->overflow_wait, &__wait,
					TASK_UNINTERRUPTIBLE);
			spin_unlock_irq(&bitmap->lock);
J
Jens Axboe 已提交
1383
			io_schedule();
1384 1385 1386 1387
			finish_wait(&bitmap->overflow_wait, &__wait);
			continue;
		}

1388
		switch (*bmc) {
1389 1390
		case 0:
			bitmap_file_set_bit(bitmap, offset);
1391
			bitmap_count_page(bitmap, offset, 1);
1392 1393 1394 1395
			/* fall through */
		case 1:
			*bmc = 2;
		}
1396

1397 1398 1399 1400 1401 1402 1403
		(*bmc)++;

		spin_unlock_irq(&bitmap->lock);

		offset += blocks;
		if (sectors > blocks)
			sectors -= blocks;
1404 1405
		else
			sectors = 0;
1406 1407 1408
	}
	return 0;
}
1409
EXPORT_SYMBOL(bitmap_startwrite);
1410 1411

void bitmap_endwrite(struct bitmap *bitmap, sector_t offset, unsigned long sectors,
1412
		     int success, int behind)
1413
{
1414 1415
	if (!bitmap)
		return;
1416
	if (behind) {
1417 1418
		if (atomic_dec_and_test(&bitmap->behind_writes))
			wake_up(&bitmap->behind_wait);
1419 1420 1421
		pr_debug("dec write-behind count %d/%lu\n",
			 atomic_read(&bitmap->behind_writes),
			 bitmap->mddev->bitmap_info.max_write_behind);
1422
	}
1423 1424 1425
	if (bitmap->mddev->degraded)
		/* Never clear bits or update events_cleared when degraded */
		success = 0;
1426

1427
	while (sectors) {
N
NeilBrown 已提交
1428
		sector_t blocks;
1429 1430 1431 1432 1433 1434 1435 1436 1437 1438
		unsigned long flags;
		bitmap_counter_t *bmc;

		spin_lock_irqsave(&bitmap->lock, flags);
		bmc = bitmap_get_counter(bitmap, offset, &blocks, 0);
		if (!bmc) {
			spin_unlock_irqrestore(&bitmap->lock, flags);
			return;
		}

1439 1440 1441 1442
		if (success &&
		    bitmap->events_cleared < bitmap->mddev->events) {
			bitmap->events_cleared = bitmap->mddev->events;
			bitmap->need_sync = 1;
1443
			sysfs_notify_dirent_safe(bitmap->sysfs_can_clear);
1444 1445
		}

1446
		if (!success && !NEEDED(*bmc))
1447 1448
			*bmc |= NEEDED_MASK;

1449
		if (COUNTER(*bmc) == COUNTER_MAX)
1450 1451
			wake_up(&bitmap->overflow_wait);

1452
		(*bmc)--;
1453
		if (*bmc <= 2) {
1454
			set_page_attr(bitmap,
1455 1456 1457
				      filemap_get_page(
					      bitmap,
					      offset >> CHUNK_BLOCK_SHIFT(bitmap)),
1458
				      BITMAP_PAGE_PENDING);
1459 1460
			bitmap->allclean = 0;
		}
1461 1462 1463 1464
		spin_unlock_irqrestore(&bitmap->lock, flags);
		offset += blocks;
		if (sectors > blocks)
			sectors -= blocks;
1465 1466
		else
			sectors = 0;
1467 1468
	}
}
1469
EXPORT_SYMBOL(bitmap_endwrite);
1470

N
NeilBrown 已提交
1471
static int __bitmap_start_sync(struct bitmap *bitmap, sector_t offset, sector_t *blocks,
1472
			       int degraded)
1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488
{
	bitmap_counter_t *bmc;
	int rv;
	if (bitmap == NULL) {/* FIXME or bitmap set as 'failed' */
		*blocks = 1024;
		return 1; /* always resync if no bitmap */
	}
	spin_lock_irq(&bitmap->lock);
	bmc = bitmap_get_counter(bitmap, offset, blocks, 0);
	rv = 0;
	if (bmc) {
		/* locked */
		if (RESYNC(*bmc))
			rv = 1;
		else if (NEEDED(*bmc)) {
			rv = 1;
1489 1490 1491 1492
			if (!degraded) { /* don't set/clear bits if degraded */
				*bmc |= RESYNC_MASK;
				*bmc &= ~NEEDED_MASK;
			}
1493 1494 1495 1496 1497 1498
		}
	}
	spin_unlock_irq(&bitmap->lock);
	return rv;
}

N
NeilBrown 已提交
1499
int bitmap_start_sync(struct bitmap *bitmap, sector_t offset, sector_t *blocks,
1500 1501 1502 1503 1504 1505 1506 1507 1508 1509
		      int degraded)
{
	/* bitmap_start_sync must always report on multiples of whole
	 * pages, otherwise resync (which is very PAGE_SIZE based) will
	 * get confused.
	 * So call __bitmap_start_sync repeatedly (if needed) until
	 * At least PAGE_SIZE>>9 blocks are covered.
	 * Return the 'or' of the result.
	 */
	int rv = 0;
N
NeilBrown 已提交
1510
	sector_t blocks1;
1511 1512 1513 1514 1515 1516 1517 1518 1519 1520

	*blocks = 0;
	while (*blocks < (PAGE_SIZE>>9)) {
		rv |= __bitmap_start_sync(bitmap, offset,
					  &blocks1, degraded);
		offset += blocks1;
		*blocks += blocks1;
	}
	return rv;
}
1521
EXPORT_SYMBOL(bitmap_start_sync);
1522

N
NeilBrown 已提交
1523
void bitmap_end_sync(struct bitmap *bitmap, sector_t offset, sector_t *blocks, int aborted)
1524 1525 1526
{
	bitmap_counter_t *bmc;
	unsigned long flags;
1527 1528

	if (bitmap == NULL) {
1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542
		*blocks = 1024;
		return;
	}
	spin_lock_irqsave(&bitmap->lock, flags);
	bmc = bitmap_get_counter(bitmap, offset, blocks, 0);
	if (bmc == NULL)
		goto unlock;
	/* locked */
	if (RESYNC(*bmc)) {
		*bmc &= ~RESYNC_MASK;

		if (!NEEDED(*bmc) && aborted)
			*bmc |= NEEDED_MASK;
		else {
1543
			if (*bmc <= 2) {
1544 1545
				set_page_attr(bitmap,
					      filemap_get_page(bitmap, offset >> CHUNK_BLOCK_SHIFT(bitmap)),
1546
					      BITMAP_PAGE_PENDING);
1547 1548
				bitmap->allclean = 0;
			}
1549 1550 1551 1552 1553
		}
	}
 unlock:
	spin_unlock_irqrestore(&bitmap->lock, flags);
}
1554
EXPORT_SYMBOL(bitmap_end_sync);
1555 1556 1557 1558 1559 1560 1561 1562

void bitmap_close_sync(struct bitmap *bitmap)
{
	/* Sync has finished, and any bitmap chunks that weren't synced
	 * properly have been aborted.  It remains to us to clear the
	 * RESYNC bit wherever it is still on
	 */
	sector_t sector = 0;
N
NeilBrown 已提交
1563
	sector_t blocks;
N
NeilBrown 已提交
1564 1565
	if (!bitmap)
		return;
1566 1567
	while (sector < bitmap->mddev->resync_max_sectors) {
		bitmap_end_sync(bitmap, sector, &blocks, 0);
N
NeilBrown 已提交
1568 1569 1570
		sector += blocks;
	}
}
1571
EXPORT_SYMBOL(bitmap_close_sync);
N
NeilBrown 已提交
1572 1573 1574 1575

void bitmap_cond_end_sync(struct bitmap *bitmap, sector_t sector)
{
	sector_t s = 0;
N
NeilBrown 已提交
1576
	sector_t blocks;
N
NeilBrown 已提交
1577 1578 1579 1580 1581 1582 1583 1584

	if (!bitmap)
		return;
	if (sector == 0) {
		bitmap->last_end_sync = jiffies;
		return;
	}
	if (time_before(jiffies, (bitmap->last_end_sync
1585
				  + bitmap->mddev->bitmap_info.daemon_sleep)))
N
NeilBrown 已提交
1586 1587 1588 1589
		return;
	wait_event(bitmap->mddev->recovery_wait,
		   atomic_read(&bitmap->mddev->recovery_active) == 0);

1590
	bitmap->mddev->curr_resync_completed = sector;
1591
	set_bit(MD_CHANGE_CLEAN, &bitmap->mddev->flags);
N
NeilBrown 已提交
1592 1593 1594 1595 1596
	sector &= ~((1ULL << CHUNK_BLOCK_SHIFT(bitmap)) - 1);
	s = 0;
	while (s < sector && s < bitmap->mddev->resync_max_sectors) {
		bitmap_end_sync(bitmap, s, &blocks, 0);
		s += blocks;
1597
	}
N
NeilBrown 已提交
1598
	bitmap->last_end_sync = jiffies;
1599
	sysfs_notify(&bitmap->mddev->kobj, NULL, "sync_completed");
1600
}
1601
EXPORT_SYMBOL(bitmap_cond_end_sync);
1602

1603
static void bitmap_set_memory_bits(struct bitmap *bitmap, sector_t offset, int needed)
1604 1605
{
	/* For each chunk covered by any of these sectors, set the
1606
	 * counter to 1 and set resync_needed.  They should all
1607 1608
	 * be 0 at this point
	 */
1609

N
NeilBrown 已提交
1610
	sector_t secs;
1611 1612 1613 1614
	bitmap_counter_t *bmc;
	spin_lock_irq(&bitmap->lock);
	bmc = bitmap_get_counter(bitmap, offset, &secs, 1);
	if (!bmc) {
1615
		spin_unlock_irq(&bitmap->lock);
1616
		return;
1617
	}
1618
	if (!*bmc) {
1619
		struct page *page;
1620
		*bmc = 1 | (needed ? NEEDED_MASK : 0);
1621 1622
		bitmap_count_page(bitmap, offset, 1);
		page = filemap_get_page(bitmap, offset >> CHUNK_BLOCK_SHIFT(bitmap));
1623
		set_page_attr(bitmap, page, BITMAP_PAGE_PENDING);
1624
		bitmap->allclean = 0;
1625 1626
	}
	spin_unlock_irq(&bitmap->lock);
1627 1628
}

1629 1630 1631 1632 1633 1634
/* dirty the memory and file bits for bitmap chunks "s" to "e" */
void bitmap_dirty_bits(struct bitmap *bitmap, unsigned long s, unsigned long e)
{
	unsigned long chunk;

	for (chunk = s; chunk <= e; chunk++) {
1635
		sector_t sec = (sector_t)chunk << CHUNK_BLOCK_SHIFT(bitmap);
1636 1637
		bitmap_set_memory_bits(bitmap, sec, 1);
		bitmap_file_set_bit(bitmap, sec);
1638 1639 1640 1641 1642 1643
		if (sec < bitmap->mddev->recovery_cp)
			/* We are asserting that the array is dirty,
			 * so move the recovery_cp address back so
			 * that it is obvious that it is dirty
			 */
			bitmap->mddev->recovery_cp = sec;
1644 1645 1646
	}
}

1647 1648 1649 1650 1651 1652
/*
 * flush out any pending updates
 */
void bitmap_flush(mddev_t *mddev)
{
	struct bitmap *bitmap = mddev->bitmap;
1653
	long sleep;
1654 1655 1656 1657 1658 1659 1660

	if (!bitmap) /* there was no bitmap */
		return;

	/* run the daemon_work three time to ensure everything is flushed
	 * that can be
	 */
1661
	sleep = mddev->bitmap_info.daemon_sleep * 2;
1662
	bitmap->daemon_lastrun -= sleep;
1663
	bitmap_daemon_work(mddev);
1664
	bitmap->daemon_lastrun -= sleep;
1665
	bitmap_daemon_work(mddev);
1666
	bitmap->daemon_lastrun -= sleep;
1667
	bitmap_daemon_work(mddev);
1668 1669 1670
	bitmap_update_sb(bitmap);
}

1671 1672 1673
/*
 * free memory that was allocated
 */
1674
static void bitmap_free(struct bitmap *bitmap)
1675 1676 1677 1678 1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 1694 1695 1696
{
	unsigned long k, pages;
	struct bitmap_page *bp;

	if (!bitmap) /* there was no bitmap */
		return;

	/* release the bitmap file and kill the daemon */
	bitmap_file_put(bitmap);

	bp = bitmap->bp;
	pages = bitmap->pages;

	/* free all allocated memory */

	if (bp) /* deallocate the page memory */
		for (k = 0; k < pages; k++)
			if (bp[k].map && !bp[k].hijacked)
				kfree(bp[k].map);
	kfree(bp);
	kfree(bitmap);
}
1697

1698 1699 1700 1701 1702 1703 1704
void bitmap_destroy(mddev_t *mddev)
{
	struct bitmap *bitmap = mddev->bitmap;

	if (!bitmap) /* there was no bitmap */
		return;

1705
	mutex_lock(&mddev->bitmap_info.mutex);
1706
	mddev->bitmap = NULL; /* disconnect from the md device */
1707
	mutex_unlock(&mddev->bitmap_info.mutex);
1708 1709
	if (mddev->thread)
		mddev->thread->timeout = MAX_SCHEDULE_TIMEOUT;
1710

1711 1712 1713
	if (bitmap->sysfs_can_clear)
		sysfs_put(bitmap->sysfs_can_clear);

1714 1715
	bitmap_free(bitmap);
}
1716 1717 1718 1719 1720 1721 1722 1723

/*
 * initialize the bitmap structure
 * if this returns an error, bitmap_destroy must be called to do clean up
 */
int bitmap_create(mddev_t *mddev)
{
	struct bitmap *bitmap;
1724
	sector_t blocks = mddev->resync_max_sectors;
1725 1726
	unsigned long chunks;
	unsigned long pages;
1727
	struct file *file = mddev->bitmap_info.file;
1728
	int err;
1729
	struct sysfs_dirent *bm = NULL;
1730

A
Alexey Dobriyan 已提交
1731
	BUILD_BUG_ON(sizeof(bitmap_super_t) != 256);
1732

1733
	if (!file
1734
	    && !mddev->bitmap_info.offset) /* bitmap disabled, nothing to do */
1735 1736
		return 0;

1737
	BUG_ON(file && mddev->bitmap_info.offset);
1738

1739
	bitmap = kzalloc(sizeof(*bitmap), GFP_KERNEL);
1740 1741 1742 1743
	if (!bitmap)
		return -ENOMEM;

	spin_lock_init(&bitmap->lock);
1744 1745
	atomic_set(&bitmap->pending_writes, 0);
	init_waitqueue_head(&bitmap->write_wait);
1746
	init_waitqueue_head(&bitmap->overflow_wait);
1747
	init_waitqueue_head(&bitmap->behind_wait);
1748

1749 1750
	bitmap->mddev = mddev;

1751 1752
	if (mddev->kobj.sd)
		bm = sysfs_get_dirent(mddev->kobj.sd, NULL, "bitmap");
1753
	if (bm) {
1754
		bitmap->sysfs_can_clear = sysfs_get_dirent(bm, NULL, "can_clear");
1755 1756 1757 1758
		sysfs_put(bm);
	} else
		bitmap->sysfs_can_clear = NULL;

1759
	bitmap->file = file;
1760 1761
	if (file) {
		get_file(file);
1762 1763 1764 1765
		/* As future accesses to this file will use bmap,
		 * and bypass the page cache, we must sync the file
		 * first.
		 */
1766
		vfs_fsync(file, 1);
1767
	}
1768
	/* read superblock from bitmap file (this sets mddev->bitmap_info.chunksize) */
1769 1770 1771 1772 1773 1774 1775 1776 1777 1778
	if (!mddev->bitmap_info.external) {
		/*
		 * If 'MD_ARRAY_FIRST_USE' is set, then device-mapper is
		 * instructing us to create a new on-disk bitmap instance.
		 */
		if (test_and_clear_bit(MD_ARRAY_FIRST_USE, &mddev->flags))
			err = bitmap_new_disk_sb(bitmap);
		else
			err = bitmap_read_sb(bitmap);
	} else {
1779 1780 1781 1782 1783 1784 1785
		err = 0;
		if (mddev->bitmap_info.chunksize == 0 ||
		    mddev->bitmap_info.daemon_sleep == 0)
			/* chunksize and time_base need to be
			 * set first. */
			err = -EINVAL;
	}
1786
	if (err)
1787
		goto error;
1788

1789
	bitmap->daemon_lastrun = jiffies;
1790
	bitmap->chunkshift = ffz(~mddev->bitmap_info.chunksize);
1791 1792

	/* now that chunksize and chunkshift are set, we can use these macros */
1793
	chunks = (blocks + CHUNK_BLOCK_RATIO(bitmap) - 1) >>
1794
			CHUNK_BLOCK_SHIFT(bitmap);
1795
	pages = (chunks + PAGE_COUNTER_RATIO - 1) / PAGE_COUNTER_RATIO;
1796 1797 1798 1799 1800 1801 1802

	BUG_ON(!pages);

	bitmap->chunks = chunks;
	bitmap->pages = pages;
	bitmap->missing_pages = pages;

1803
#ifdef INJECT_FATAL_FAULT_1
1804 1805
	bitmap->bp = NULL;
#else
1806
	bitmap->bp = kzalloc(pages * sizeof(*bitmap->bp), GFP_KERNEL);
1807
#endif
1808
	err = -ENOMEM;
1809
	if (!bitmap->bp)
1810
		goto error;
1811

1812 1813 1814 1815 1816 1817 1818 1819 1820 1821 1822 1823 1824 1825 1826 1827
	printk(KERN_INFO "created bitmap (%lu pages) for device %s\n",
		pages, bmname(bitmap));

	mddev->bitmap = bitmap;


	return (bitmap->flags & BITMAP_WRITE_ERROR) ? -EIO : 0;

 error:
	bitmap_free(bitmap);
	return err;
}

int bitmap_load(mddev_t *mddev)
{
	int err = 0;
1828
	sector_t start = 0;
1829 1830 1831 1832 1833 1834 1835 1836 1837 1838 1839 1840
	sector_t sector = 0;
	struct bitmap *bitmap = mddev->bitmap;

	if (!bitmap)
		goto out;

	/* Clear out old bitmap info first:  Either there is none, or we
	 * are resuming after someone else has possibly changed things,
	 * so we should forget old cached info.
	 * All chunks should be clean, but some might need_sync.
	 */
	while (sector < mddev->resync_max_sectors) {
N
NeilBrown 已提交
1841
		sector_t blocks;
1842 1843 1844 1845 1846
		bitmap_start_sync(bitmap, sector, &blocks, 0);
		sector += blocks;
	}
	bitmap_close_sync(bitmap);

1847 1848 1849 1850 1851 1852 1853 1854
	if (mddev->degraded == 0
	    || bitmap->events_cleared == mddev->events)
		/* no need to keep dirty bits to optimise a
		 * re-add of a missing device */
		start = mddev->recovery_cp;

	err = bitmap_init_from_disk(bitmap, start);

1855
	if (err)
1856
		goto out;
1857

1858
	mddev->thread->timeout = mddev->bitmap_info.daemon_sleep;
1859
	md_wakeup_thread(mddev->thread);
1860

1861 1862
	bitmap_update_sb(bitmap);

1863 1864 1865
	if (bitmap->flags & BITMAP_WRITE_ERROR)
		err = -EIO;
out:
1866
	return err;
1867
}
1868
EXPORT_SYMBOL_GPL(bitmap_load);
1869

1870 1871 1872 1873
static ssize_t
location_show(mddev_t *mddev, char *page)
{
	ssize_t len;
1874
	if (mddev->bitmap_info.file)
1875
		len = sprintf(page, "file");
1876
	else if (mddev->bitmap_info.offset)
1877
		len = sprintf(page, "%+lld", (long long)mddev->bitmap_info.offset);
1878
	else
1879 1880 1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 1903 1904 1905 1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929
		len = sprintf(page, "none");
	len += sprintf(page+len, "\n");
	return len;
}

static ssize_t
location_store(mddev_t *mddev, const char *buf, size_t len)
{

	if (mddev->pers) {
		if (!mddev->pers->quiesce)
			return -EBUSY;
		if (mddev->recovery || mddev->sync_thread)
			return -EBUSY;
	}

	if (mddev->bitmap || mddev->bitmap_info.file ||
	    mddev->bitmap_info.offset) {
		/* bitmap already configured.  Only option is to clear it */
		if (strncmp(buf, "none", 4) != 0)
			return -EBUSY;
		if (mddev->pers) {
			mddev->pers->quiesce(mddev, 1);
			bitmap_destroy(mddev);
			mddev->pers->quiesce(mddev, 0);
		}
		mddev->bitmap_info.offset = 0;
		if (mddev->bitmap_info.file) {
			struct file *f = mddev->bitmap_info.file;
			mddev->bitmap_info.file = NULL;
			restore_bitmap_write_access(f);
			fput(f);
		}
	} else {
		/* No bitmap, OK to set a location */
		long long offset;
		if (strncmp(buf, "none", 4) == 0)
			/* nothing to be done */;
		else if (strncmp(buf, "file:", 5) == 0) {
			/* Not supported yet */
			return -EINVAL;
		} else {
			int rv;
			if (buf[0] == '+')
				rv = strict_strtoll(buf+1, 10, &offset);
			else
				rv = strict_strtoll(buf, 10, &offset);
			if (rv)
				return rv;
			if (offset == 0)
				return -EINVAL;
1930 1931
			if (mddev->bitmap_info.external == 0 &&
			    mddev->major_version == 0 &&
1932 1933 1934 1935 1936 1937 1938 1939 1940 1941 1942 1943 1944 1945 1946 1947 1948 1949 1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966
			    offset != mddev->bitmap_info.default_offset)
				return -EINVAL;
			mddev->bitmap_info.offset = offset;
			if (mddev->pers) {
				mddev->pers->quiesce(mddev, 1);
				rv = bitmap_create(mddev);
				if (rv) {
					bitmap_destroy(mddev);
					mddev->bitmap_info.offset = 0;
				}
				mddev->pers->quiesce(mddev, 0);
				if (rv)
					return rv;
			}
		}
	}
	if (!mddev->external) {
		/* Ensure new bitmap info is stored in
		 * metadata promptly.
		 */
		set_bit(MD_CHANGE_DEVS, &mddev->flags);
		md_wakeup_thread(mddev->thread);
	}
	return len;
}

static struct md_sysfs_entry bitmap_location =
__ATTR(location, S_IRUGO|S_IWUSR, location_show, location_store);

static ssize_t
timeout_show(mddev_t *mddev, char *page)
{
	ssize_t len;
	unsigned long secs = mddev->bitmap_info.daemon_sleep / HZ;
	unsigned long jifs = mddev->bitmap_info.daemon_sleep % HZ;
1967

1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040 2041 2042 2043 2044 2045 2046 2047 2048 2049 2050 2051 2052 2053 2054 2055 2056 2057 2058 2059
	len = sprintf(page, "%lu", secs);
	if (jifs)
		len += sprintf(page+len, ".%03u", jiffies_to_msecs(jifs));
	len += sprintf(page+len, "\n");
	return len;
}

static ssize_t
timeout_store(mddev_t *mddev, const char *buf, size_t len)
{
	/* timeout can be set at any time */
	unsigned long timeout;
	int rv = strict_strtoul_scaled(buf, &timeout, 4);
	if (rv)
		return rv;

	/* just to make sure we don't overflow... */
	if (timeout >= LONG_MAX / HZ)
		return -EINVAL;

	timeout = timeout * HZ / 10000;

	if (timeout >= MAX_SCHEDULE_TIMEOUT)
		timeout = MAX_SCHEDULE_TIMEOUT-1;
	if (timeout < 1)
		timeout = 1;
	mddev->bitmap_info.daemon_sleep = timeout;
	if (mddev->thread) {
		/* if thread->timeout is MAX_SCHEDULE_TIMEOUT, then
		 * the bitmap is all clean and we don't need to
		 * adjust the timeout right now
		 */
		if (mddev->thread->timeout < MAX_SCHEDULE_TIMEOUT) {
			mddev->thread->timeout = timeout;
			md_wakeup_thread(mddev->thread);
		}
	}
	return len;
}

static struct md_sysfs_entry bitmap_timeout =
__ATTR(time_base, S_IRUGO|S_IWUSR, timeout_show, timeout_store);

static ssize_t
backlog_show(mddev_t *mddev, char *page)
{
	return sprintf(page, "%lu\n", mddev->bitmap_info.max_write_behind);
}

static ssize_t
backlog_store(mddev_t *mddev, const char *buf, size_t len)
{
	unsigned long backlog;
	int rv = strict_strtoul(buf, 10, &backlog);
	if (rv)
		return rv;
	if (backlog > COUNTER_MAX)
		return -EINVAL;
	mddev->bitmap_info.max_write_behind = backlog;
	return len;
}

static struct md_sysfs_entry bitmap_backlog =
__ATTR(backlog, S_IRUGO|S_IWUSR, backlog_show, backlog_store);

static ssize_t
chunksize_show(mddev_t *mddev, char *page)
{
	return sprintf(page, "%lu\n", mddev->bitmap_info.chunksize);
}

static ssize_t
chunksize_store(mddev_t *mddev, const char *buf, size_t len)
{
	/* Can only be changed when no bitmap is active */
	int rv;
	unsigned long csize;
	if (mddev->bitmap)
		return -EBUSY;
	rv = strict_strtoul(buf, 10, &csize);
	if (rv)
		return rv;
	if (csize < 512 ||
	    !is_power_of_2(csize))
		return -EINVAL;
	mddev->bitmap_info.chunksize = csize;
	return len;
}

static struct md_sysfs_entry bitmap_chunksize =
__ATTR(chunksize, S_IRUGO|S_IWUSR, chunksize_show, chunksize_store);

2060 2061 2062 2063 2064 2065 2066 2067 2068 2069 2070 2071 2072 2073 2074 2075 2076 2077 2078 2079 2080 2081 2082 2083 2084 2085 2086 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097 2098 2099 2100 2101 2102 2103 2104 2105 2106 2107 2108 2109 2110 2111 2112
static ssize_t metadata_show(mddev_t *mddev, char *page)
{
	return sprintf(page, "%s\n", (mddev->bitmap_info.external
				      ? "external" : "internal"));
}

static ssize_t metadata_store(mddev_t *mddev, const char *buf, size_t len)
{
	if (mddev->bitmap ||
	    mddev->bitmap_info.file ||
	    mddev->bitmap_info.offset)
		return -EBUSY;
	if (strncmp(buf, "external", 8) == 0)
		mddev->bitmap_info.external = 1;
	else if (strncmp(buf, "internal", 8) == 0)
		mddev->bitmap_info.external = 0;
	else
		return -EINVAL;
	return len;
}

static struct md_sysfs_entry bitmap_metadata =
__ATTR(metadata, S_IRUGO|S_IWUSR, metadata_show, metadata_store);

static ssize_t can_clear_show(mddev_t *mddev, char *page)
{
	int len;
	if (mddev->bitmap)
		len = sprintf(page, "%s\n", (mddev->bitmap->need_sync ?
					     "false" : "true"));
	else
		len = sprintf(page, "\n");
	return len;
}

static ssize_t can_clear_store(mddev_t *mddev, const char *buf, size_t len)
{
	if (mddev->bitmap == NULL)
		return -ENOENT;
	if (strncmp(buf, "false", 5) == 0)
		mddev->bitmap->need_sync = 1;
	else if (strncmp(buf, "true", 4) == 0) {
		if (mddev->degraded)
			return -EBUSY;
		mddev->bitmap->need_sync = 0;
	} else
		return -EINVAL;
	return len;
}

static struct md_sysfs_entry bitmap_can_clear =
__ATTR(can_clear, S_IRUGO|S_IWUSR, can_clear_show, can_clear_store);

2113 2114 2115 2116 2117 2118 2119 2120 2121 2122 2123 2124 2125 2126 2127 2128 2129 2130 2131 2132 2133
static ssize_t
behind_writes_used_show(mddev_t *mddev, char *page)
{
	if (mddev->bitmap == NULL)
		return sprintf(page, "0\n");
	return sprintf(page, "%lu\n",
		       mddev->bitmap->behind_writes_used);
}

static ssize_t
behind_writes_used_reset(mddev_t *mddev, const char *buf, size_t len)
{
	if (mddev->bitmap)
		mddev->bitmap->behind_writes_used = 0;
	return len;
}

static struct md_sysfs_entry max_backlog_used =
__ATTR(max_backlog_used, S_IRUGO | S_IWUSR,
       behind_writes_used_show, behind_writes_used_reset);

2134 2135 2136 2137 2138
static struct attribute *md_bitmap_attrs[] = {
	&bitmap_location.attr,
	&bitmap_timeout.attr,
	&bitmap_backlog.attr,
	&bitmap_chunksize.attr,
2139 2140
	&bitmap_metadata.attr,
	&bitmap_can_clear.attr,
2141
	&max_backlog_used.attr,
2142 2143 2144 2145 2146 2147 2148
	NULL
};
struct attribute_group md_bitmap_group = {
	.name = "bitmap",
	.attrs = md_bitmap_attrs,
};